Relationships among tree-species composition, vegetation structure, and forest breeding birds in southern Illinois

Oaks (Quercus spp.) have dominated eastern forests of the United States for centuries; however, current disturbance regimes discourage oak recruitment and allow shade-tolerant mesophytic species (e.g., maples, Acer spp.) to out-compete oaks. I assessed the effects of mesophication on bird communities by examining differences in breeding bird community structure, abundance, and diversity across 8 and 12 deciduous forest stands in southern Illinois during 2009 and 2010, respectively, using line transects, and by examining a 5-year monitoring data set from across the Shawnee National Forest in 2005-2009. I predicted that variation in bird community structure between maple- and oak-dominated forests can be explained by differential availability of foraging niches. Forest stands used in 2009-2010 were separated along a gradient of hard-mast tree composition, which was defined as the percentage of tree basal area in the stand contributed by oaks and hickories (Carya spp.). Linear regression and Akaike's Information Criteria were used to assess habitat-association models for 7 bird community metrics: bird species diversity, species richness, overall abundance, and abundance of aerial foragers, bark gleaners, foliage gleaners, and ground gleaners. Bird species diversity (Shannon-Wiener H') and species richness ranged from 2.97 to 3.15, and 29 to 37, respectively, over both years. Bird species diversity and species richness were best modeled by a negative relationship with % hardmast tree basal area across both years, whereas overall abundance was best modeled by a positive relationship with understory woody stem diversity. Detection rates for foraging guilds were best modeled by various metrics of habitat structure. Aerial foragers, bark gleaners, foliage gleaners, and ground gleaners responded positively to stem density, downed coarse woody debris density, basal area, and stem diversity, respectively. I used non-metric multidimensional scaling (NMDS) and analysis of similarity (ANOSIM) to examine the degree of dissimilarity among bird communities and site type. In 2010, the bird community differed overall, with communities in oak forests tending (P = 0.09-0.11) to differ those in non-oak and mixed-mesophytic sites. Analysis of a 5-year data set yielded similar results. All models tested for bird-species diversity and species richness were competing, suggesting no individual habitat factor was a strong predictor. Overall abundance and abundance of aerial foragers, bark gleaners and foliage gleaners showed negative relationships to hardmast basal area in all years combined. Ground gleaners responded positively to tree diversity. A post-hoc analysis revealed that overall bird abundance and abundance of foliage-, bark- and ground-gleaning guilds responded positively to an index of riparian areas. Resource use during the breeding season may be shifted to mesic habitats, possibly due to increased resource availability in terms of arthropods and water. Results indicated that mesophication may not have the predicted effects on forest-breeding bird communities, and that vegetation structure was more important in determining bird community structure than tree composition in small-scale forest stands during the breeding season. Resident and over-wintering bird species may be most affected by the loss of oaks due to use during time periods when mesic habitats do not supplement resources. Managers should consider maintenance of a diversity of forest types to maximize avian diversity.